SNMP traps capture system monitoring information that is sent as an asynchronous notification from the SNMP agent (storage system) to the SNMP manager. There are three types of SNMP traps: standard, built-in, and user-defined.
A trap can be used to periodically check for different operational thresholds or failures, which are defined in the MIB. If a threshold is reached or failure is detected, the SNMP agent (storage system) sends a message (trap) to the traphosts alerting them of the event.
- Standard SNMP traps
- These traps are defined in RFC 1215. There are five standard SNMP traps that are supported by Data ONTAP: coldStart, warmStart, linkDown, linkUp, and authenticationFailure.
- Built-in SNMP traps
- Built-in traps are predefined in Data ONTAP and are automatically sent to the network management stations on the traphost list if an event occurs. These traps are defined in the custom MIB, such as diskFailedShutdown, cpuTooBusy, and volumeNearlyFull.
Each built-in trap is identified by a unique trap code.
- User-defined SNMP traps
- User-defined traps are defined by the snmp traps command. These traps are sent using proxy trap IDs 11 through 18, which correspond to a trap's MIB priority.